By default, the following are installed on a typical installation of Akamai App Platform: Gitea, Tekton, Argo CD, the platform API, and the App Platform Console. The Console is the self-service UI and uses the platform API to generate validated configuration code. This configuration code is then committed to Gitea (in the values repository), which will trigger the pre-configured Tekton pipeline to apply the changes.

In some cases, you might not want to use the Console and the platform API, but instead install and manage configuration of the platform using a custom pipeline. Possible use cases for this scenario include:

Pushing configuration changes to multiple clusters at the same time (edge use-case).
Not allowing configuration changes to be made by developers.
Only using the open source apl-core project.

Example Helm chart

The example Helm chart values below shows how to install the platform with a team, 2 workloads and 2 services:

cluster:
  name: ${clusterName}
  provider: ${clusterProvider}
otomi::
  adminPassword: ${password}
apps:
  argocd:
    enabled: true
teamConfig:
  demo:
    password: ${somePassword}
    id: myapp
    networkPolicy:
      egressPublic: true
      ingressPrivate: true
    services:
      - name: front-end
        ingressClassName: platform
        networkPolicy:
            ingressPrivate:
                mode: DenyAll
        ownHost: true
        port: 8080
        type: public
        headers:
          response:
            set:
              - name: Cache-Control
                value: no-store
              - name: Pragma
                value: no-cache
      - name: api
        networkPolicy:
            ingressPrivate:
                mode: AllowOnly
                allow:
                  team: myapp
                  service: front-end
        port: 8081
    workloads:
      - name: front-end
        path: ksvc
        revision: v1.0.1
        selectedChart: ksvc
        url: https://github.com/redkubes/apl-charts.git
      - name: api
        path: deployment
        revision: v1.0.1
        selectedChart: deployment
        url: https://github.com/redkubes/apl-charts.git
files:
  env/teams/workloads/demo/front-end.yaml: |
    values: |
        fullnameOverride: front-end
        image:
            repository: harbor.${domainSuffix}/team-demo/front-end
            tag: latest
        containerPorts:
            - name: http1
            containerPort: 8080
            protocol: TCP
        resources:
            requests:
                cpu: 200m
                memory: 32Mi
        autoscaling:
            minReplicas: 0
            maxReplicas: 10    
  env/teams/workloads/demo/api.yaml: |
    values: |
      image:
        image:
            repository: harbor.${domainSuffix}/team-demo/api
            tag: latest
        containerPorts:
            - name: http1
            containerPort: 8080
            protocol: TCP
        resources:
            requests:
                cpu: 200m
                memory: 32Mi
        autoscaling:
            minReplicas: 1
            maxReplicas: 10

Customize installation

You can now make changes in this configuration and apply them directly to the cluster:

Install using your custom values:
```
helm install -f values.yaml apl apl/apl
```
After installation:
- All required ingress resources will be created.
- Istio (including the virtual services for public exposed services with HTTP response headers) will be installed and configured.
- Network policies.
- Argo CD and Argo CD applications and application sets are created to deploy the front-end and api workloads.
Change the values

You can extend the values with all supported configuration options.
Uninstall the release:
```
helm uninstall apl --no-hooks
```
Re-install the chart with the adjusted values:
```
helm install -f values.yaml apl apl/apl
```